COMPONENT 5. METHODOLOGIES FOR IMPROVING MEASUREMENT OF ALCOHOL CONSUMPTION AND[unreadable] PROBLEMS (All years). The aims of the proposed component are to: 1) Improve estimates of the mean and[unreadable] sources of variation in drink size and ethanol content for beer, wine and spirits drinks consumed at home through[unreadable] beaker size measurement and brand identification of subjects' usual drinks. 2) Estimate mean values and explore[unreadable] the sources of variation in on-premise beer, wine and spirits-based drinks. Bartender focus groups will explore[unreadable] perceptions of drink size and sources of variation by those who make the drinks. Results will be used to inform[unreadable] the design of intensive direct drink measurement in a small sample of bars and restaurants where a variety of[unreadable] drinks will be purchased on multiple visits. Drink size will be directly measured and the percentage of ethanol in[unreadable] each beverage will be determined by brand for beer, wine and straight spirits and will be measured using an[unreadable] Analox(R) Alcohol Analyzer for mixed spirits drinks. A 10-county random sample of 80 establishments of multiple[unreadable] types will be used to estimate the mean size and analyze key sources of variation in nine representative beer,[unreadable] wine and spirits drinks. 3) Improve estimates of ethanol intake measures from beer, wine, spirits and all[unreadable] beverages combined through the application of empirical drink ethanol content mean and variance estimates to[unreadable] NAS11 based on demographics and drink-specific questions in the survey. 4) Demonstrate the effect of[unreadable] Interactive Voice Response (IVR) technology when used during a telephone interview (in NAS11) on sensitive[unreadable] measures of alcohol use, alcohol-related problems, and risky behavior. 5) Within the NAS11 survey, the[unreadable] relationships between multiple measures of the same construct across time frames and question formats will be[unreadable] quantified and the sources and importance of differences will be analyzed in predictive models. 6) Exploratory[unreadable] analyses will compare alcohol measures in the NAS with those in three major survey series; The National[unreadable] Epidemiologic Survey of Alcohol and Related Conditions (NESARC), the Behavioral Risk Factor Surveillance[unreadable] System (BRFSS) and the National Household Survey of Drug Use and Health (NSDUH). These analyses will[unreadable] consider not only overall and demographic sub-group means and prevalence estimates but also the full[unreadable] distribution of responses.[unreadable] The ARG Center will continue to play a central role in the development, improvement and evaluation of[unreadable] methodologies for measuring diverse aspects of alcohol use and a wide range of alcohol-related problems and[unreadable] related health-risk behaviors. Results will be disseminated to researchers, practitioners and the public.[unreadable]